Bloomberg L.p. v. Securities and Exchange Commission

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued January 10, 2022 Decided August 16, 2022
No. 21-1088
BLOOMBERG L.P.,
PETITIONER
v.
SECURITIES AND EXCHANGE COMMISSION,
RESPONDENT
FINANCIAL INDUSTRY REGULATORY AUTHORITY,
INTERVENOR
On Petition for Review of an Order
of the Securities and Exchange Commission
Keith Bradley argued the cause for petitioner. With him
on the briefs was ScheLeese Goudy.
Thomas A. Burns was on the brief for amicus curiae
Healthy Markets Association in support of petitioner.
Theodore J. Weiman, Senior Litigation Counsel,
Securities and Exchange Commission, argued the cause for
respondent. With him on the brief were Michael A. Conley,
Solicitor, and Tracey A. Hardin, Assistant General Counsel.

-- 1 of 27 --

2
Nowell D. Bamberger was on the brief for intervenor
Financial Industry Regulatory Authority, Inc. in support of
respondent. Giovanni P. Prezioso entered an appearance.
Paul S. Mishkin and David B. Toscano were on the brief
for amici curiae Intercontinental Exchange, Inc., et al. in
support of respondent.
Before: WILKINS, KATSAS and JACKSON, Circuit Judges.
Opinion for the Court filed by Circuit Judge WILKINS.
WILKINS, Circuit Judge: Petitioner Bloomberg L.P.
(“Bloomberg”) seeks review of the Securities and Exchange
Commission’s (the “Commission” or “SEC”) decision to
approve new reporting requirements proposed by the Financial
Industry Regulatory Authority, Inc. (“FINRA”), Intervenor-
for-Respondent, affecting underwriter members in the
corporate bond market. FINRA represented to the SEC that
market inefficiencies in the corporate bond market reduce
market participation, decrease liquidity, and increase
transaction and opportunity costs. To address these problems,
FINRA proposed to consolidate and provide market-wide
access to “core” reference data for new issues of corporate
bonds through a subscription-based service. The Commission
ultimately concluded that FINRA’s proposal would impose a
limited burden on competition and enable market participants
to obtain broad, uniform access to corporate bond reference
data before the first transaction in a new-issue bond.
Accordingly, the Commission approved FINRA’s proposal.
 Circuit Judge, now Justice, Jackson was a member of the panel at
the time the case was argued but did not participate in this opinion.

-- 2 of 27 --

3
Importantly, though, during the rulemaking process,
various commenters raised concerns about FINRA’s proposed
data service. In relevant part, Bloomberg commented that
FINRA did not provide any information about how much it will
cost to build and maintain the database, and to what extent
FINRA will pass those costs along to market participants.
For the reasons explained below, we find that pursuant to
the Administrative Procedure Act, 5 U.S.C. § 706(2)(A), the
Commission’s approval of FINRA’s proposal was arbitrary
and capricious because the Commission neglected to give a
reasoned explanation in response to Bloomberg’s significant
concerns about the costs that FINRA, as well as market
participants, will incur in connection to the creation and
maintenance of the data service. Accordingly, we grant
Bloomberg’s petition for review on the grounds that the
Commission’s failure to respond to significant public
comments about the costs associated with FINRA’s proposal
was arbitrary and capricious. We deny Bloomberg’s petition
for review with respect to its remaining arguments. We remand
without vacatur for the Commission to respond appropriately.
I.
FINRA is a private association of securities broker-dealers
that regulates the conduct of broker-dealers—people or firms
in the business of buying or selling securities on behalf of
customers, their own accounts or both—and formulates and
enforces standards for trading and brokerage. Because of its
status as a self-regulatory organization, Section 19(b)(2)(C)(i)
of the Securities and Exchange Act of 1934 (the “Exchange
Act” or “Act”), 15 U.S.C. § 78s(b)(2)(C)(i), imposes on
FINRA “a duty to promulgate and enforce rules governing the
conduct of [its] members, under the oversight of the SEC.” See
NetCoalition v. SEC, 715 F.3d 342, 344–45 (D.C. Cir. 2013)

-- 3 of 27 --

4
(“NetCoalition II”) (internal quotation marks and citations
omitted). The Act provides that FINRA must submit any
proposed rules or rule changes to the SEC, which “shall
approve” proposals if it finds them “consistent with the
requirements of” the Exchange Act and applicable SEC rules
and regulations. 15 U.S.C. § 78s(b)(2)(C)(i).
Bloomberg is a global business and financial information
company. Among other things, it is a private data vendor that
dominates the market for private data services. See Pet’r’s
Opening Br. at 8. Bloomberg obtains newly issued bond
reference data from underwriters—usually before other data
vendors and market participants—and then sells that data to
market participants.
Reference data is very important. When corporations issue
bonds, the underwriters who buy and sell those bonds to market
participants compile certain reference data—details and terms,
such as the name of the issuer and identification numbers—and
provide them to data vendors like Bloomberg. Without
reference data, trading platforms cannot list a bond for trading.
Timing matters, too. The hours and days that immediately
follow a bond offering are generally a highly active trading
period. Accordingly, market participants, namely investors,
seek out timely reference data about new corporate bonds
coming to market from vendors like Bloomberg. According to
FINRA, because each reference data provider collects and
distributes new issue reference data “from different sources and
at different speeds[,]” the reference data is not necessarily
“consistent, timely and accurate across reference data
providers.” J.A. 1–2.

-- 4 of 27 --

5
II.
In 2017, the SEC established the Fixed Income Market
Structure Advisory Committee (“FIMSAC” or “Advisory
Committee”) to offer advice and recommendations to the
Commission on the structure of the fixed income market.
(Corporate bonds are considered fixed income securities.)
FIMSAC is composed of individuals representing a range of
perspectives on the fixed income markets, including corporate
bond investors, broker-dealers, underwriters, academics, and
data vendors. FIMSAC’s Technology and Electronic Trading
Subcommittee (the “Subcommittee”), which represents a
cross-section of fixed income market participants, determined
that there are gaps in the corporate bond reference data market,
both in terms of when data vendors make data available and the
data’s accuracy. The Subcommittee attributed these disparities
to (1) the fact that private data vendors—like Bloomberg—are
not obligated to provide impartial access to new issue reference
data; (2) vendors lacking equal access to information from
underwriters; and (3) the resulting confusion, which increases
transaction costs and impedes competition in the corporate
bond market.
The Subcommittee met several times over the course of
seven months to fill the gaps in the corporate bond reference
data market. It ultimately recommended that FINRA, as an
impartial entity, establish a consolidated new issue reference
data service that makes core reference data available to all
subscribers in a timely and commercially reasonable manner.
The Subcommittee’s recommendation paralleled another
proposal the SEC had previously approved. In 2008, the SEC
accepted the Municipal Securities Rulemaking Board’s
proposal to establish a centralized, self-regulatory
organization-mandated data service that would provide

-- 5 of 27 --

6
participants in the municipal bond market with timely access to
“core” reference data and streamline the trade of municipal
bonds. The petition before us concerns a proposal for a similar
service in the corporate bond market.
FIMSAC unanimously endorsed the Subcommittee’s
recommendation. In October 2018, FIMSAC recommended
that FINRA establish a corporate bond new issue reference data
service—similar to the municipal bond market service that the
SEC approved in 2008—to mitigate the disparities in the
market for corporate new issue reference data.
A.
To that end, FINRA performed an Economic Impact
Assessment: an analysis of the corporate bond market. FINRA
coordinated independent outreach to eleven participants in the
corporate bond market: four data vendors, three underwriters,
two trading platforms, and two clearing firms. Insights from
these entities led FINRA to discover the same problems
FIMSAC did: data vendors receive reference data through
different channels at different times, which leads to untimely,
inconsistent, and inaccurate distribution of corporate bond new
issue reference data to market participants. Thereafter, FINRA
set out to develop a solution that would address gaps in the
availability of accurate, complete, and timely access to
corporate bond new issue reference data.
FINRA developed a proposal to establish a centralized
new issue reference data service for corporate bonds. First,
FINRA would require its member underwriters to report to
FINRA—before trading on a new issue begins—32 unique data
elements that are integral to the valuation, trade, and settlement
of corporate bonds. Collectively, these elements constitute
“core” reference data. Next, FINRA would create a database

-- 6 of 27 --

7
to consolidate the data and provide market-wide access through
a fee subscription.
B.
On March 27, 2019, pursuant to Section 19(b)(1) of the
Exchange Act, 15 U.S.C. § 78s(b)(1), FINRA proposed to the
SEC the establishment of a consolidated fee-based data service
that would provide market-wide access to new issue reference
data.
FINRA Rule 6760 “requires certain data elements—those
sufficient to identify the security accurately—to be reported
before the execution of the first transaction, and all remaining
data elements to be reported within 15 minutes of the Time of
Execution of the first transaction.” J.A. 2 n.4. Rule 6760(b)
currently requires underwriters to provide certain new issue
reference data to FINRA. FINRA’s proposal would amend
Rule 6760(b) to require underwriters to report additional data
elements to FINRA. In relevant part, the amended version of
Rule 6760(b) would require underwriters to report the
following core information:
(A) The International Securities Identification
Number (ISIN); (B) the currency; (C) the issue
date; (D) the first settle date; (E) the interest
accrual date; (F) the day count description; (G)
the coupon frequency; (H) the first coupon
payment date; (I) a Regulation S indicator; (J)
the security type; (K) the bond type; (L) the first
coupon period type; (M) a convertible indicator;
(N) a call indicator; (O) the first call date; (P) a
put indicator; (Q) the first put date; (R) the
minimum increment; (S) the minimum
piece/denomination; (T) the issuance amount;

-- 7 of 27 --

8
(U) the first call price; (V) the first put price;
(W) the coupon type; (X) rating (TRACE
Grade); (Y) a perpetual maturity indicator; (Z)
a Payment-In-Kind (PIK) indicator; (AA) first
conversion date; (BB) first conversion ratio;
(CC) spread; (DD) reference rate; (EE) floor;
and (FF) underlying entity ticker.
J.A. 281. FINRA represented to the SEC that it would submit
a separate filing to establish fees for the service and would
implement its proposed service if those fees are adopted.1 Id.
C.
In April 2019, the Commission published notice of
FINRA’s proposed rule change in the Federal Register. See
Notice of Filing of a Proposed Rule Change to Establish a
Corporate Bond New Issue Reference Data Service, 84 Fed.
Reg. 13,977 (Apr. 8, 2019). On July 1, 2019, the Commission
instituted proceedings under Section 19(b)(2)(B) of the
Exchange Act, 15 U.S.C. § 78s(b)(2)(B), to determine whether
to approve the rule change.
1 Initially, in its March 2019 proposal, FINRA proposed that it would
make the new-issue data available to subscribers for a fee of $250
per month (for internal use) or $6,000 per month (to retransmit the
data to other users). On October 3, 2019, FINRA filed Partial
Amendment No. 1, which was withdrawn because of a non-
substantive administrative error. That same day, FINRA filed Partial
Amendment No. 2, in which it removed the fee component and
tweaked some of the core data fields required from underwriters.
FINRA withdrew the proposed subscription fees because
commenters urged FINRA to provide more information to justify the
fees. Upon withdrawal, FINRA stated that it would further evaluate
the appropriate fee structure.

-- 8 of 27 --

9
The Commission received 30 comments in response to
FINRA’s proposal, including five from Bloomberg, one from
FINRA, and one from FIMSAC. Supporters of FINRA’s
proposal, including Intercontinental Exchange, Inc. (“ICE”)
Bonds, ICE Data Services, and Charles River Development, all
of whom are market participants, commented “that currently
there is no uniform, universally available mechanism for
providing market participants with consistent and timely access
to reference data about corporate bonds on the day a newly
issued corporate bond commences trading,” J.A. 193, and
explained that timely access to reference data is crucial for the
valuation, trade, and settlement of corporate bonds. Supporters
of FINRA’s proposal also pointed to another issue with the
current market for reference points: reference data is not
available to all market participants before the trading of a new
issue, which puts certain market participants at a competitive
disadvantage. Ultimately, supporters of FINRA’s proposal
contended that a FINRA-operated centralized data reporting
requirement for new corporate bond issues would increase the
efficiency of the corporate bond market and reduce trading and
research costs.
Opponents of FINRA’s proposal, including Bloomberg,
the Heritage Foundation, and the Healthy Markets Association,
submitted comments arguing that FINRA failed to establish the
existence of a market structure problem that requires regulatory
intervention, as mandated under Section 15A(b)(6) of the
Exchange Act, 15 U.S.C. § 78o–3(b)(6). Assuming that a
problem exists, Bloomberg suggested that it was “questionable
whether a single [self-regulatory organization] would provide
more accurate, complete and timely service than competing
private sector providers” and FINRA provided no evidence that
its proposal would reduce broken trades and errors. J.A. 193
n.28 (internal quotation marks and citation omitted). Further,
Bloomberg submitted that the impact of any errors in a

-- 9 of 27 --

10
centralized system would be magnified. Id. The U.S. Chamber
of Commerce’s Center for Capital Markets Competitiveness
commented that FINRA’s proposal would increase regulatory
and liability burdens for underwriters without any clear benefit.
Bloomberg echoed this concern, contending that the rule’s
imposition of additional burdens on underwriters would
disproportionately impact smaller underwriters.
In addition, Bloomberg commented that FINRA’s
proposal was “antithetical to the most foundational principles
of administrative law and cost-benefit analysis” because
FINRA failed to quantify the direct and indirect costs of its
proposed service (or explain why certain costs could not be
quantified). J.A. 162. Bloomberg cautioned that self-
regulatory organizations and agencies should not be “permitted
to ignore regulatory burdens in this manner.” Id. Otherwise,
“agencies could propose laudable programs heedless of their
pricetags, seek their provisional approval without respect to
cost, and then—once established—propose a fee that was by
now necessary to sustain an already approved program.” Id.
“[A]t a minimum,” Bloomberg commented, “the Commission
must . . . know what [the] costs [of the proposed rule] are.” See
id.
On December 4, 2019, the SEC’s Division of Trading and
Markets (the “Division”), which is authorized by the SEC and
federal regulations to approve proposed rule changes, accepted
FINRA’s proposal and concluded that it was consistent with
Sections 15A(b)(6) and 15A(b)(9) of the Act. See 15 U.S.C.
§§ 78o–3(b)(6), (9).
On December 18, 2019, Bloomberg petitioned the
Commission for review of the Division’s December 2019
approval order. The Commission conducted a de novo review

-- 10 of 27 --

11
of whether FINRA’s proposal is consistent with the Exchange
Act and applicable rules and regulations.
III.
Section 15A(b) of the Exchange Act, 15 U.S.C. § 78o–
3(b), lays out numerous requirements the SEC must consider
when reviewing a rule change. Section 15A(b)(5) requires that
“[t]he rules of the association provide for the equitable
allocation of reasonable dues, fees, and other charges among
members and issuers and other persons using any facility or
system which the association operates or controls.” Id. § 78o–
3(b)(5). Section 15A(b)(6) provides that FINRA’s rules must
be “designed” “to promote just and equitable principles of
trade”; “to foster cooperation and coordination with persons
engaged in regulating, clearing, settling, processing
information with respect to, and facilitating transactions in
securities”; “to remove impediments to and perfect the
mechanism of a free and open market and a national market
system”; and “in general, to protect investors and the public
interest.” Id. § 78o–3(b)(6). Finally, Section 15A(b)(9) states
that FINRA’s rules may not “impose any burden on
competition not necessary or appropriate.” Id. § 78o–3(b)(9).
As a general matter, when the SEC is engaged in the
review of a rule of a self-regulatory organization like FINRA,
Section 3(f) of the Exchange Act provides that the SEC must
consider “whether the action will promote efficiency,
competition, and capital formation.” Id. § 78c(f). As
mentioned, the SEC “shall approve” a rule proposal if it is
“consistent with the requirements of” the Exchange Act and
applicable SEC rules and regulations. Id. § 78s(b)(2)(C)(i). It
is ultimately FINRA’s burden to demonstrate that its proposed
rule change is consistent with the Act and applicable rules and
regulations. 17 C.F.R. § 201.700(b)(3).

-- 11 of 27 --

12
A.
On January 15, 2021, the Commission upheld the
Division’s approval order, finding FINRA’s proposed rule
change consistent with the Exchange Act, namely Section 3(f),
15 U.S.C. § 78c(f), Section 15A(b)(6), id. § 78o–3(b)(6), and
Section 15A(b)(9), id. § 78o–3(b)(9).
In its order, the Commission declared that “there is an
inefficiency in the collection and availability of reference data
for newly issued corporate bonds.” J.A. 281 (footnote
omitted). “[T]his inefficiency results in an information
asymmetry in the market for newly issued corporate bond
reference data that can disadvantage many market
participants.” Id. Further, “[t]his information asymmetry
inhibits these market participants from transacting in the
secondary market for newly issued bonds. . . .” Id. The
Commission concluded that “FINRA’s proposal is reasonably
designed to address this information asymmetry to the benefit
of the marketplace” by “mak[ing] certain reference data
available to market participants in a timely, accessible, and
impartial manner.” Id.
The Commission also addressed several issues raised by
commenters, which included, among other things, (1) “whether
information asymmetry [actually] exists in the current
marketplace”; (2) “the proposal’s burden on underwriters”; (3)
“the proposal’s effect on competition among reference data
vendors”; and (4) “the lack of information regarding fees” for
FINRA’s data service. Id. at 282.
First, the Commission reasoned that the record—namely
comments provided by reference data providers at the October
2018 FIMSAC meeting and FINRA’s outreach to market

-- 12 of 27 --

13
participants through its Economic Impact Assessment—
establishes the existence of a reference data access problem.
FIMSAC’s meeting transcript provides that Bloomberg
“typically has the timeliest access to newly issued bond
reference data on the first day a bond trades, as it enjoys the
voluntary cooperation of underwriters,” id. at 286 n.91, while
other data vendors face challenges with collecting and
distributing reference data. The meeting transcript also reflects
that Bloomberg’s dominance is attributable to underwriters’
unwillingness to distribute reference data to all market
participants because widespread distribution increases the risk
of inaccuracies. Accordingly, data vendors “must expend
substantial time and effort gathering information from multiple
sources . . . ultimately resulting in an unnecessary market
inefficiency.” Id. at 286–87. This hinders timely market-wide
participation in corporate bond trading.
Next, the Commission assuaged concerns about the burden
of additional reporting requirements on underwriters. The
Commission observed that the impact of FINRA’s proposal on
how underwriters currently distribute data or how data vendors
conduct business is uncertain, but underwriters “already
have . . . data reporting processes in place and have incurred
the costs of establishing those processes.” Id. at 298.
Accordingly, underwriters would have the choice to continue
providing new issue reference data to data vendors as they see
fit.
The Commission then turned to the proposal’s general
burden on competition. The Commission conceded that “the
impact on competition is uncertain,” but nevertheless
concluded that “any burden on competition would both be
limited and justified by the evidence in the record
demonstrating an information asymmetry that can
disadvantage many market participants due to the lack of

-- 13 of 27 --

14
timely access to basic information that is important for the
identification, valuation and settlement of newly issued
corporate bonds.” Id. at 282. Further, FINRA’s proposal
would not require market participants to purchase data from
FINRA. The Commission noted that even Bloomberg
acknowledged “that market participants currently demand
more reference data fields than FINRA is proposing to collect.”
Id. at 298 n.249.
Finally, the Commission addressed concerns about fees
associated with FINRA’s proposal as well as the overall cost of
the proposed service. The Commission first noted that FINRA
cannot charge fees for its data service without the
Commission’s authorization, and the Commission would not
approve FINRA’s fee proposal absent a showing that it
complies with the Exchange Act and the SEC’s rules regarding
proposed fee changes. Moreover, proposed fee changes are
subject to public notice and comment. Even if FINRA’s fee
proposal becomes effective upon filing with the Commission,
the Commission has authority to suspend the fee and order
proceedings to assess whether the fee proposal is consistent
with the Exchange Act.
Further, the Commission disagreed with commenters—
including Bloomberg—who suggested that the agency could
not adequately assess the proposal’s economic effects without
knowing the fees FINRA would charge and the costs FINRA
would incur in connection to building the service. In response
to such comments, the Commission reasoned that FINRA’s
later “fee filing . . . will merit a consideration of FINRA’s cost
to build the New Issue Reference Data Service.” J.A. 301.
“[T]he costs of the system, which will be better known once
the system is built, will be necessary to assess whether FINRA
has proposed a fee for that service that is consistent with the
Act, including Section 15A(b)(5).” Id. All in all, the

-- 14 of 27 --

15
Commission reasoned that FINRA’s proposal is justified
because it will resolve a widespread corporate bond reference
data problem to the benefit of the market and market
participants.
B.
On March 15, 2021, Bloomberg filed a timely petition for
this Court to review the Commission’s January 2021 order.
Bloomberg contends that the Commission’s decision should be
vacated as arbitrary and capricious and unsupported by
substantial evidence pursuant to the Administrative Procedure
Act, 5 U.S.C. § 706(2)(A), because (1) there is insufficient
evidence of an actual reference data access problem in the
corporate bond market; (2) FINRA’s data service will impede
competition in the corporate bond market by imposing undue
burdens on underwriters and ousting incumbent data vendors,
in violation of Sections 3(f) and 15A(b)(9) of the Exchange
Act, 15 U.S.C. §§ 78c(f), 78o–3(b)(9); (3) the Commission
failed to conduct a cost-benefit analysis of FINRA’s proposal,
in violation of its obligation under Section 3(f) of the Exchange
Act, id. § 78c(f); and (4) the Commission improperly allowed
FINRA to evade review of the fees associated with its data
service, thereby ignoring its obligation under Section
15A(b)(5) of the Act, id. § 78o–3(b)(5).
For the reasons explained below, we find that the
Commission’s approval of FINRA’s proposal was arbitrary
and capricious under the Administrative Procedure Act, 5
U.S.C. § 706(2)(A), because the Commission failed to respond
to significant and relevant concerns Bloomberg raised in its
comments objecting to FINRA’s proposal. Specifically, the
agency did not provide a reasoned response to Bloomberg’s
comments that FINRA failed to quantify the direct and indirect
costs of its proposed data service (or explain why certain costs

-- 15 of 27 --

16
could not be quantified), and failed to explain how the costs
incurred for building the service will be paid if the Commission
disapproves FINRA’s proposed fee structure in subsequent
proceedings. As such, the Commission’s decision to approve
FINRA’s proposal was not the product of reasoned
decisionmaking.
IV.
This Court has jurisdiction to review the Commission’s
January 2021 order pursuant to Section 25(a) of the Exchange
Act, 15 U.S.C. § 78y(a). We review the order under the
Administrative Procedure Act, which requires us to hold
unlawful agency action that is “arbitrary, capricious, an abuse
of discretion, or otherwise not in accordance with law” or that
is “unsupported by substantial evidence.” 5 U.S.C. §§
706(2)(A), (E); Susquehanna Int’l Grp., LLP v. SEC, 866 F.3d
442, 445 (D.C. Cir. 2017) (internal quotation marks omitted).
“To satisfy the ‘arbitrary and capricious’ standard, ‘the agency
must examine the relevant data and articulate a satisfactory
explanation for its action including a rational connection
between the facts found and the choice made.’” Susquehanna,
866 F.3d at 445 (quoting Motor Vehicle Mfrs. Ass’n v. State
Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983)).
A.
At the outset, Bloomberg contends that certain documents
in the Joint Appendix—namely the transcript and other
materials from an October 2018 FIMSAC meeting—are not a
part of the official administrative record in this case because
the SEC did not list these materials as a separate exhibit in the
certified index of the record. Pet’r’s Opening Br. at 29–30;
Pet’r’s Reply Br. at 5–6.

-- 16 of 27 --

17
Bloomberg is right that the SEC neglected to include
certain materials in connection to the October 2018 FIMSAC
meeting in the administrative record. But the SEC’s apparent
error does not justify vacatur of the Commission’s January
2021 order. To be sure, federal courts base their review of
agency action on the record before the agency. See 5 U.S.C. §
706. But, in this instance, the SEC’s “omission” of certain
FIMSAC documents “is of little consequence, and certainly
does not warrant vacatur” because the “parties did include the
[omitted materials] in the Joint Appendix, making [them]
available for our perusal” and referred to them throughout their
briefs. MD Pharm., Inc. v. Drug. Enf. Admin., 133 F.3d 8, 15
(D.C. Cir. 1998) (“The Rules allow parties to supply materials
that were improperly omitted from the record.”); see also FED.
R. APP. P. 16(b). In short, we will not vacate the order because
of the SEC’s omission or disregard the omitted record material
in our resolution of this petition.
B.
We turn next to Bloomberg’s substantive challenges to the
Commission’s January 2021 order.
First, Bloomberg contends that FINRA—and by
extension, the SEC—relied on “anonymous sources,
corroborated by complaints from competing data vendors that
they struggle to get data from underwriters.” Pet’r’s Opening
Br. at 15–16. Bloomberg characterizes FINRA’s outreach
effort as biased, resulting in “anecdotes [that]
cannot . . . constitute substantial evidence” of reference data
access issues. See id. at 16. Rather than discovering evidence
that traders or investors lack access to reference data,
Bloomberg says FINRA’s “process was simply a mechanism
for Bloomberg’s rivals to complain about their challenges.” Id.
at 30 (emphasis in original). Further, Bloomberg invokes

-- 17 of 27 --

18
Susquehanna to bolster its claim that the SEC “uncritically
accepted” FINRA’s evidence. Id. (citing Susquehanna, 866
F.3d at 448).
In Susquehanna, a clearing agency proposed to make
changes to the fees that exchanges pay to the agency. 866 F.3d
at 443–44. The SEC accepted the clearing agency’s capital
plan proposal. The SEC conceded in Susquehanna that it relied
on an assessment performed by the clearing agency’s
consultant; it did not undertake its own analysis. Id. at 446–47.
The Court criticized the SEC’s “unquestioning reliance” on the
clearing agency’s defense of its own actions and the SEC’s
failure to “critically review[] [the clearing agency’s] analysis
or perform[] its own [analysis].” Id. at 447. The Court
concluded that the SEC failed to ascertain whether the dividend
level proposed by the clearing agency was reasonable. Id.
Instead, “the SEC took [the clearing agency’s] word for it.” Id.
For that reason, the Court determined that the SEC’s approval
of the clearing agency’s proposal to change exchange fees was
arbitrary and capricious. Id. at 449.
The case before us is nothing like Susquehanna. First, it
was FIMSAC—an SEC-backed advisory committee—that
independently discovered the existence of a reference data
access problem in the market for new issue corporate bond
reference data. FIMSAC is comprised of 23 diverse members,
including investors, broker-dealers, underwriters, the SEC’s
former chief economist, and data vendors; these participants
unanimously agreed that participants in the corporate bond
market lack timely, uniform, and accurate access to new issue
reference data. Moreover, FIMSAC’s recommendation to
create a solution preceded FINRA’s undertaking of an
Economic Impact Assessment and market participant outreach.
The January 2021 order reflects that the SEC found FINRA’s
analysis consistent with other evidence, including materials

-- 18 of 27 --

19
from the October 2018 FIMSAC meeting. Additionally,
contrary to Bloomberg’s assertions, the participants in the
FIMSAC proceedings are not anonymous; many are identified
on the SEC’s website in FIMSAC’s public meeting transcripts
and subcommittee meeting minutes.
All in all, by suggesting that the Commission’s decision
rests on speculation that a reference data access problem exists,
Bloomberg overlooks substantial evidence submitted by
FINRA—not merely anecdotes from supposedly biased market
participants—that there are information asymmetries and
inefficiencies in the market for new issue corporate bond
reference data. For all these reasons, Bloomberg’s argument—
that the SEC merely took FINRA’s word for the existence of a
reference data access problem—does not hold water.
C.
The parties also dispute whether FINRA’s proposal
offends Section 15A(b)(9) of the Exchange Act, which
provides that FINRA’s rules should “not impose any burden on
competition not necessary or appropriate in furtherance of the
purposes of” the Exchange Act, 15 U.S.C. § 78o–3(b)(9), and
Section 3(f), which requires the SEC to “consider . . . whether
the action will promote efficiency, competition, and capital
formation” in its review of FINRA’s proposed rule.
Id. § 78c(f).
Bloomberg contends that FINRA’s data service constitutes
“government-compelled data collection” that is inconsistent
with Section 15A(b)(9), 15 U.S.C. § 78o–3(b)(9), because it
will burden competition by “nationaliz[ing] corporate-bond
reference data, replacing multiple private enterprises with a
single regulatory utility.” Pet’r’s Opening Br. at 16, 31. In
Bloomberg’s view, the Commission downplayed the risk of

-- 19 of 27 --

20
creating a “quasi-governmental monopoly [that] will replace
competitive private providers,” which rendered the
Commission’s January 2021 order arbitrary and capricious.
See id. at 22. Further, Bloomberg says FINRA’s proposed
amendment to Rule 6760(b) will require underwriters to
provide data to the FINRA-mandated database, thereby
disincentivizing underwriters from giving data to private
vendors. See id. at 21–22. Finally, Bloomberg suggests that
implementation of FINRA’s proposed rule change could render
data service vendors unviable. Id. at 34.
In its response brief, the SEC contends that the
Commission’s approval of FINRA’s proposal is consistent
with Section 15A(b)(9), 15 U.S.C. § 78o–3(b)(9), and Section
3(f), id. § 78c(f), because any limited burden on competition is
outweighed “by the benefits of alleviating the information
asymmetry that disadvantages many market participants.”
Resp.’s Br. at 34. Moreover, the SEC reiterates that FINRA’s
proposal will promote market participation and suggests that
the Exchange Act “does not limit its focus on competition to
one narrow sector of the market; it focuses on the impact of
competition in the market as a whole.” Id. Furthermore, the
SEC does not share Bloomberg’s concern that FINRA’s data
service—which will only require underwriters to report core
reference data—will oust private data vendors like Bloomberg.
Id. at 37. Finally, the SEC points out that FIMSAC’s
recommendation to establish a new issue reference data service
was based on the success of a similarly centralized, self-
regulatory organization-mandated service in the municipal
bond market, where competition continues to exist. See id. at
17. For these reasons, the SEC contends that FINRA’s data
service will foster competition, not burden it.
The SEC’s argument is compelling, and it accurately
encapsulates the Commission’s requirements under the

-- 20 of 27 --

21
Exchange Act. The Exchange Act obligates the Commission
to consider the effects of proposed rules and regulations on the
market as a whole, not just narrow sections of the market. For
example, in Business Roundtable v. Securities & Exchange
Commission, this Court emphasized the Commission’s “unique
obligation to consider the effect of a new rule upon ‘efficiency,
competition, and capital formation’” as well as its economic
consequences. 647 F.3d 1144, 1148 (D.C. Cir. 2011) (quoting
15 U.S.C. § 78c(f)). Failure to fulfill this obligation “makes
promulgation of the rule arbitrary and capricious and not in
accordance with law.” Id.; see also Chamber of Com. v. SEC,
412 F.3d 133, 143 (D.C. Cir. 2005) (“the Commission [has a]
statutory obligation to determine as best as it can the economic
implications of the rule it has proposed”); New York Stock
Exch. LLC v. SEC, 962 F.3d 541, 559 (D.C. Cir. 2020).
All in all, Bloomberg’s argument that FINRA’s data
service will impose an unnecessary burden on competition
lacks merit. First, Bloomberg mischaracterizes FINRA’s
proposed data service as a competitive endeavor that will
displace incumbent data vendors, and not a supplementary
service that will foster competition and improve efficiency and
timeliness in the reference data market. Also, Bloomberg
overlooks the existence of underwriters’ existing data reporting
processes, which make continued data collection “less
burdensome than if new processes had to be established.”
Resp.’s Br. at 40.
Moreover, Bloomberg’s own comments during the agency
proceedings undermine its arguments here. As the
Commission mentioned in its January 2021 order, Bloomberg
acknowledged that “market participants currently demand
more reference data fields than FINRA is proposing to collect.”
J.A. 298 n.249. In other words, FINRA’s proposal will
facilitate additional competition by imposing data reporting

-- 21 of 27 --

22
requirements on underwriters, which will in turn level the
playing field for private data providers who face a reference
data access problem. Underwriters would still be free to
provide a broader scope of reference data elements to data
vendors, and in turn, data vendors can meet the market’s
demand for more comprehensive data.
To be sure, the Commission conceded in its January 2021
order that “the impact [of FINRA’s proposal] on competition
is uncertain” and dependent on market participants’ response.
J.A. 282. Nevertheless, given the narrow scope of FINRA’s
proposed amendment to Rule 6760(b) and significant evidence
of a potentially positive impact on competition in the corporate
bond market, including input from diverse market participants
with expertise in the corporate bond market, we hold that the
Commission concluded appropriately that “the limited set of
data proposed to be reported and disseminated to allow for the
identification, valuation and settlement of new issue corporate
bonds is unlikely to supplant the demand for a more
comprehensive reference database.” J.A. 298–99. At its core,
the Administrative Procedure Act’s substantial evidence
standard “requires more than a scintilla, but can be satisfied by
something less than a preponderance of the evidence.” Epsilon
Elecs., Inc. v. U.S. Dep’t of Treasury, 857 F.3d 913, 925 (D.C.
Cir. 2017) (internal quotation marks and citations omitted). “If
that threshold is met, we must uphold the agency’s judgment
regarding the relevant facts. . . .” Id.
In sum, on the record before us, we hold that the
Commission had substantial evidence to support its
determination that FINRA’s proposal is consistent with
Sections 3(f) and 15A(b)(9) of the Exchange Act, 15 U.S.C.
§§ 78c(f), 78o–3(b)(9), because it will impose only a limited,
justifiable burden on competition.

-- 22 of 27 --

23
D.
We next address Bloomberg’s final two arguments—that
the Commission failed to weigh the proposed rule’s costs and
benefits, in violation of Section 3(f) of the Exchange Act, 15
U.S.C. § 78c(f), and arbitrarily allowed FINRA to evade
review of the fee component of its data service, in violation of
Section 15A(b)(5) of the Exchange Act, which requires the
Commission to assess whether FINRA’s proposed rule
“provide[s] for the equitable allocation of reasonable . . . fees.”
15 U.S.C. § 78o–3(b)(5).
First, we reject Bloomberg’s argument that the
Commission arbitrarily allowed FINRA to evade review of the
fee component of its data service, in violation of Section
15A(b)(5) of the Exchange Act. See id. As mentioned,
FINRA’s proposed fee is subject to public notice and comment
and the Commission retains the authority to suspend FINRA’s
fee-based service and order proceedings to assess whether the
fee proposal is consistent with the Exchange Act. In its order,
the Commission acknowledged its obligation to ensure that
FINRA’s pending fee proposal is consistent with the Exchange
Act. See J.A. 301. In sum, the SEC will consider fees in a
future rulemaking, and the agency reasonably concluded that it
has control over those fees and will not approve them unless
they are reasonable and equitable such that they are consistent
with the Exchange Act.
We turn next to Bloomberg’s argument that the
Commission’s decision to approve FINRA’s proposal was
arbitrary and capricious because the Commission failed to
consider the costs FINRA will incur in building the system and
the extent to which FINRA will pass along those costs to
market participants, especially if FINRA’s data service is

-- 23 of 27 --

24
unreasonably expensive. See Pet’r’s Opening Br. at 42. In
Bloomberg’s view, Section 3(f) of the Exchange Act, 15 U.S.C.
§ 78c(f), obligated the Commission to evaluate the costs and
benefits of FINRA’s proposed data service, but the
Commission abdicated this responsibility. See Pet’r’s Opening
Br. at 35 (citing Chamber of Com., 412 F.3d at 143, and Bus.
Roundtable, 647 F.3d at 1149).
Although Bloomberg has raised the issue of whether
Section 3(f) of the Exchange Act, 15 U.S.C. § 78c(f), imposed
on the Commission a requirement to include costs incurred by
FINRA in a cost-benefit analysis of FINRA’s proposed rule,
we need not reach that question because the Commission’s
failure to respond adequately to Bloomberg’s comments
rendered its decision arbitrary and capricious under the
Administrative Procedure Act. Bloomberg raised relevant
concerns about the direct and indirect costs of FINRA’s
proposal, but the Commission brushed them aside. As such,
the Commission’s decision was not the product of reasoned
decisionmaking.
“Under the APA, whenever agencies promulgate a rule
that intends to create new law, rights, or duties . . . they must
engage in a process known as notice-and-comment
rulemaking.” Oakbrook Land Holdings, LLC v. Comm’r of
Internal Revenue, 28 F.4th 700, 710 (6th Cir. 2022) (cleaned
up); see also 5 U.S.C. § 553(b). Pursuant to this process, “an
agency must respond to comments ‘that can be thought to
challenge a fundamental premise’ underlying the proposed
agency decision.” Carlson v. Postal Regul. Comm’n, 938 F.3d
337, 344 (D.C. Cir. 2019) (quoting MCI WorldCom, Inc. v.
FCC, 209 F.3d 760, 765 (D.C. Cir. 2000)); see also City of
Waukesha v. EPA, 320 F.3d 228, 257 (D.C. Cir. 2003) (“The
agency need not address every comment, but it must respond
in a reasoned manner to those that raise significant problems.”)

-- 24 of 27 --

25
(internal quotation marks and citation omitted). Indeed, “[t]he
requirement that agency action not be arbitrary or capricious
includes a requirement that the agency adequately explain its
result and respond to relevant and significant public
comments.” Pub. Citizen, Inc. v. FAA, 988 F.2d 186, 197 (D.C.
Cir. 1993) (internal quotation marks and citations omitted). In
sum, “[a]n agency’s response to public comments . . . must be
sufficient to enable the courts ‘to see what major issues of
policy were ventilated … and why the agency reacted to them
as it did.’” Carlson, 938 F.3d at 344 (quoting Del. Dep’t of
Nat. Res. & Env’tl Control v. EPA, 785 F.3d 1,17 (D.C. Cir.
2015)).
As mentioned, during notice-and-comment rulemaking,
Bloomberg “raised concerns regarding FINRA’s costs to build
and operate the new reference data service.” J.A. 301. In
response, the Commission reasoned that if FINRA
“hypothetically build[s] a New Issue Reference Data Service
at a high cost that would be unreasonable to pass on to end-
users,” then “FINRA would not be able to make a showing that
any such fees proposed to be assessed on the basis of its cost to
build the service are reasonable, as required by Section
15A(b)(5) of the Act.” Id.; 15 U.S.C. § 78o–3(b)(5). “In such
a case . . . the Commission would suspend and disapprove the
proposal.” J.A. 301.
The Commission’s analysis overlooks a key problem: if
FINRA’s data service ends up being unreasonably expensive,
then the agency cannot protect market participants from footing
the bill for it at the fees stage. To be sure, the Commission is
right that it could suspend and disapprove FINRA’s proposal
at the fees stage, see id., but at that point, FINRA will have
already incurred the financial burden of building the service.
That cost—which could be millions, or even tens of millions,
of dollars—must be paid by someone, whether the subscribers

-- 25 of 27 --

26
of the service or the broker-dealers who make up FINRA. In
short, the Commission approved FINRA’s proposal without
responding to comments that urged it to assess not only the
financial impact of the service on FINRA, but also the entities
that fund FINRA. That is not reasoned decisionmaking.
All in all, if public comments raise relevant and significant
concerns about the costs associated with a proposed rule, then
the agency should provide a reasoned response to those
comments. See, e.g., Ne. Md. Waste Disposal Auth. v. EPA,
358 F.3d 936, 950 (D.C. Cir. 2004) (per curiam) (holding that
the EPA responded adequately to members of the municipal
waste combustor industry who submitted comments
complaining about the high costs associated with regulating
pollutants produced by municipal waste combustor units). In
this case, the Commission’s failure to respond to relevant and
significant comments about the direct and indirect costs of
FINRA’s proposed data service was sufficient to render its
decision arbitrary and capricious.
Furthermore, we find that the appropriate remedy for the
agency’s error is to remand the Commission’s order approving
the proposal without vacating the order. “The decision to
vacate depends on two factors: the likelihood that
‘deficiencies’ in an order can be redressed on remand, even if
the agency reaches the same result, and the ‘disruptive
consequences’ of vacatur.” Black Oak Energy, LLC v. FERC,
725 F.3d 230, 244 (D.C. Cir. 2013) (quoting Allied-Signal v.
Nuclear Regul. Comm’n, 988 F.2d 146, 150–51 (D.C. Cir.
1993)). In this case, both factors weigh against vacatur. See
Vecinos para el Bienestar de la Comunidad Costera v. FERC,
6 F.4th 1321, 1331–32 (D.C. Cir. 2021).
First, we find that on remand, “the Commission can
redress its failure of explanation” by analyzing the costs

-- 26 of 27 --

27
FINRA will incur in building and maintaining its data service
and how the costs of building the data service will be
remunerated if the fee proposal is ultimately disapproved by
the Commission. See id. at 1332. Second, we find that vacatur
of the order would “needlessly disrupt” the Commission and
FINRA’s efforts to address market inefficiencies resulting
from untimely, inconsistent, and inaccurate collection and
dissemination of new issue reference data in the corporate bond
market. See id.
V.
In sum, we find that the Commission’s approval of
FINRA’s proposed reference data service was arbitrary and
capricious in one respect: the Commission failed to respond
adequately to Bloomberg’s concerns about the cost of building
and maintaining the program and the extent to which those
costs—which could conceivably amount to millions, or tens of
millions, of dollars—will be borne by market participants. As
such, the Commission violated the Administrative Procedure
Act and failed to engage in reasoned decisionmaking. In this
regard, Bloomberg’s petition for review is granted.
Bloomberg’s remaining arguments lack merit. Therefore,
Bloomberg’s petition for review is otherwise denied. For the
foregoing reasons, we remand to the Commission without
vacatur for further proceedings consistent with this opinion.
So ordered.

-- 27 of 27 --